Mila > Publication > Graphical Model > Active Learning for Matching Problems

Active Learning for Matching Problems

Graphical Model
Jun 2012

Active Learning for Matching Problems

Jun 2012

Effective learning of user preferences is critical to easing user burden in various types of matching problems. Equally important is active query selection to further reduce the amount of preference information users must provide. We address the problem of active learning of user preferences for matching problems, introducing a novel method for determining probabilistic matchings, and developing several new active learning strategies that are sensitive to the specific matching objective. Experiments with real-world data sets spanning diverse domains demonstrate that matching-sensitive active learning outperforms standard techniques.

Reference

Laurent Charlin, Richard Zemel, Craig Boutilier, Active Learning for Matching Problems, in: International Conference on Machine Learning (ICML), 2012

PDF

Linked Profiles